Message-Id: <sba89ec2.047@gwise-gw1.uclan.ac.uk>
Date: Wed, 19 Sep 2001 13:33:14 +0100
To: lojbab <lojbab@lojban.org>, lojban <lojban@yahoogroups.com>
Subject: Re: [lojban] META : Who is everyone (and what are they saying)
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: quoted-printable
Content-Disposition: inline
From: And Rosta <arosta@uclan.ac.uk>
Content-Length: 2344
Lines: 57

#>>> "Bob LeChevalier (lojbab)" <lojbab@lojban.org> 09/19/01 01:01am >>>
#At 06:36 PM 9/18/01 +0100, And Rosta wrote:
#>Lojbab:
#>#On one extreme we have Michael Helsem, whose poetic efforts at Lojban se=
t a
#>#very non-hardline extreme of usage.  But that extreme clearly is NOT
#>#driving usage because few users emulate Michael.  More of them try to
#>#emulate xod, or Jorge, or Nick, who each have their own styles that are
#>#more or less logically rigorous.  At the other extreme is And, who has n=
ot
#>#for the most part mirrored Michael by presenting us with a usage that
#>#reflects his image of the language.
#>
#>There are several reasons why I write so little Lojban, but there is one=
=20
#>reason
#>in particular why I don't try to influence others' usage through my own. =
This
#>reason is that there is no mechanism for abbreviation, for creating more
#>concise locutions that do not increase vagueness. There is no 'Zipf valve=
'
#>-- no mechanism for shortening locutions whose length is not appropriatel=
y
#>proportional to their frequency.
#
#The problem is that Zipfean processes usually only work once we know that=
=20
#we will be using an excessively long formulation, and until you use=20
#something a few times, we don't know what excessively long formulation=20
#needs to be Zipfed in that we can't see what the repetitive pattern is tha=
t=20
#we are trying to abbreviate.  The abbreviations we do have, like soi and=20
#sei and many of the UI like po'o, we were able to predict from natlang=20
#patterns, but you are posing new abbreviatable patterns possibly lacking=20
#parallel in the natural laanguages.

I know this, and it's clear that a pioneering loglang can't foresee everyth=
ing
that needs to be zipfed down. What I mean is that even when usage shows
what needs to be zipfed down, we then have no way to do the zipfing.
The morphology affords us no spare class of short cmavo, and the Lojban
project has not countenanced a stage at some point in the future when
zipf adjustments are made.

I have, though, seen one indication of zipfing down of lujvo:

      tilju x1 is a pedant [Adjective pedantic] (This was made by shortenin=
g the=20
      lujvo tilju'edu'u, which is perhaps the most 'proper' word for this c=
oncept)=20

http://nuzban.wiw.org/wiki/index.php?Slang%20gismu

--And.